In Japan, Expanding a Label Is What Triggers the Price Cut

Three of the largest repricings Japan confirmed this month were triggered by the same event: the drug was approved to treat another group of patients. Dupixent added COPD. Nucala added chronic rhinosinusitis with nasal polyps. Vyvdura added chronic inflammatory demyelinating polyneuropathy. Each expansion pushed sales past a benchmark, and each breach set off a cut. A fourth drug, Datroway, is being repriced on the strength of a cost-effectiveness assessment it has never undergone.

Three cuts, three label expansions, one mechanism

Sanofi’s Dupixent takes its fourth reduction under market expansion repricing, at around 3.7 percent. The trigger was COPD, added in March 2025, which pushed annual sales past ¥150bn and beyond 1.3 times the baseline benchmark, bringing the drug inside the Special Price Adjustment for Sustainable Health System and Sales Scale, known in the market as huge seller repricing.

GSK’s Nucala faces roughly 10 percent in its second market expansion repricing, after chronic rhinosinusitis with nasal polyps was added in August 2024 and sales passed ¥35bn, more than twice its benchmark. Argenx’s Vyvdura takes around 15 percent in its first, triggered by chronic inflammatory demyelinating polyneuropathy in December 2024.

Read those three together and the mechanism is visible. None was found too expensive. None failed an assessment. Each was approved to treat more people, sold more as a result, and crossed a line drawn against a forecast made before the expansion existed. The event that triggers the cut is the event every other part of the system is designed to encourage.

Datroway’s price falls on an assessment run on a different molecule

The sharper structural point sits with Daiichi Sankyo’s Datroway, facing a reduction of about 11 percent, the same figure as Gilead’s Trodelvy.

Trodelvy underwent cost-effectiveness assessment and its evaluation is recorded as complete. Datroway did not. Japan’s national register lists Datroway in the H5 category with no completed evaluation against it, which is precisely what H5 means: the price moves by reference to the pricing comparator’s result, and Trodelvy is that comparator.

The mechanism should be stated plainly. A company’s price fell on the strength of an economic evaluation conducted on a competitor’s product. There is an internal logic to it, since comparator-anchored pricing set the price to begin with, and letting the anchor move while the anchored product stays put would open an obvious gap. But it means a Japanese pricing model has to carry a variable no sponsor controls.

Against that, the FY2026 outline closes a different channel entirely. The spillover rule, which let one product’s market expansion repricing pull down therapeutically similar drugs whose own sales had not exceeded forecast, has been abolished. Anyone calling this system uniformly extractive has to account for that.

The patients who lose are the ones in the indication that never gets filed

Bullous pemphigoid, chronic inflammatory demyelinating polyneuropathy, chronic rhinosinusitis with nasal polyps. These are not abstractions in a pricing model. They are the populations that gained access because a sponsor ran the trials, filed the supplement and won the expansion.

If crossing a sales benchmark reliably produces a price cut, the commercial case for the next supplement changes. The sponsor is no longer weighing trial cost against incremental revenue, but against that revenue net of a reduction applied to the entire existing book of business. For a drug already selling well in Japan, a modest new population can be value destroying on those terms even when the clinical case is strong.

That is a different failure mode from the one usually discussed. Drug lag and drug loss describe medicines arriving late or not at all. This is harder to see, because the medicine is already there and already reimbursed. What goes missing is the indication, and nobody publishes a list of supplements that were never filed.

The system does pay for some things. Dupixent’s reduction was moderated by an orphan premium for bullous pemphigoid and a paediatric premium for bronchial asthma, so rare and paediatric populations carry protection the general case does not.

Lilly said publicly what the industry says privately

The context for all of this is Mounjaro. Japan cut its price by 25 percent under the same huge seller mechanism, effective 1 August, alongside a reduction of around 15 percent for Alnylam’s Amvuttra. On 3 August Eli Lilly criticised the decision directly, warning that unpredictable policy erodes Japan’s standing as an investment destination.

AstraZeneca Japan’s president Andy Barnett has argued along the same lines, naming market expansion repricing, the huge seller adjustment and the foreign price adjustment mechanism as the priorities for reform. Both are interested parties and should be read as such. The harder figure to dismiss is about capital rather than opinion: Japan’s share of global biopharmaceutical R&D investment fell from 12 percent to 6 percent between 2012 and 2022, on the industry association’s own account, while global spending doubled.

What the logic selects for

The FY2026 revision cut prices by an average of about 4 percent on a spending basis across roughly 15,800 products, and because Japan sits in the reference basket for other systems, a repricing triggered by Japanese volume can propagate into markets that never saw the volume.

The generalisable point is not about Japan. Any system that reprices on volume is, by construction, taxing the expansion of access, because volume and access are the same quantity counted from different ends. Japan applies that logic more mechanically than most, which makes it the clearest place to watch what the logic does. The behaviour it selects for is a drug that sells steadily to the population it launched with, and never grows.